In recent years high speed communications in excess of 10 Mbps has rapidly spread to cellular systems, broadband systems for residential environments, WiFi hotspots such as coffee shops, and so on. Some of these higher communication speeds based on single carriers systems such as CDMA, GSM, TDMA, WCDMA, etc., require a receiver architecture more advanced than a typical Rake receiver in order to function in environments with a large delay spread.
One such commonly accepted receiver architecture that is more resilient to multipath distortion is based on a Linear Minimum Mean Squared Error (LMMSE) equalizer. One of the main issues with LMMSE equalization, however, is the need for inverting relatively large matrices, which manifests itself in the form of computational and cost overhead.